Evaluation of recurrent selection in Leaming and Midland maize populations

Rates of maize (Zea mays L.) yield improvement in the United States has been reduced during the 1990's. Double-cycle breeding in a few old inbred lines and limited germplasm evaluation and utilization are possible explanations. Only one heterotic pattern (Iowa Stiff Stalk Synthetic x Lancaster Sure Crop) is being widely exploited in U.S. dent maize breeding programs. Alternative U.S. heterotic patterns have been suggested, but they received limited attention and improvement. The objectives of this research were to evaluate inbred-progeny recurrent selection in Leaming (originated in Ohio), the oldest known U.S. open-pollinated variety, and Midland Yellow Dent, an open-pollinated variety that originated in Kansas; and to evaluate the heterosis expression in the Leaming and Midland Yellow Dent cross. Response to selection was effective in both populations. Grain yield showed an average increase per cycle (P ≤ 0.01) of 0.28 Mg ha-1 (10.1%) in Leaming, and of 0.22 Mg ha-1 (14.9%) in Midland Yellow Dent. The cross between both populations had a midparent heterosis value of 4.9% before selection and of 17.8% in the third cycle of selection. The increase of grain yield heterosis (12.9%) through selection was highly significant and crosses between S1 generations showed an increased rate of inbreeding depression. Inbreeding depression due to genetic drift, however, was nonsignificant in both populations. Grain yield heterosis among improved cycles of Leaming by Midland Yellow Dent was attributed to the selection of different alleles with dominance effects in Learning. Favorable responses to selection were observed in traits that are essential for adaptation, especially in root lodging. S1-S2 recurrent selection was an effective method to adapt and improve both populations, suggesting an adequate balance between the selection among S1 and S2 progenies. Intrapopulational recurrent selection was also successful in increasing the heterotic expression between Leaming and Midland Yellow Dent. The first objective of our long-term program has been achieved and considerable improvements have been made. An alternative U.S. heterotic pattern has been confirmed and needs refinement, and an interpopulational-reciprocal recurrent selection program including improved selection cycles of Leaming and Midland Yellow Dent seems desirable.